Predictive Policing Symposium: Policy and Practice Breakout Session
Craig Uchida, Ph.D.
President, Justice & Security Strategies, Inc.
First Predictive Policing Symposium
At the first Predictive Policing Symposium, participants broke into smaller groups and reported the key elements of their
discussion back to the larger assembly.
Dr. Craig Uchida’s group discussed how hot spots analysis has become increasingly commonplace. Predictive Policing on the
other hand is still evolving. There should be less focus on violent crime and more focus on the other 90 percent of things
that police do.
Several questions were raised during the session. Can we actually predict? With the tools and technology that exist, and the
data that is available, participants felt the next level of prediction is possible. Other questions included: What predictive
elements are we looking at? Are we looking at individuals? If so, participants suggested that we “tread carefully.”
How does community policing differ? One participant said that that there does not seem to be much difference from what departments
are doing now, but perhaps predictive policing is more “high technology, high touch” because it develops relationships with
Participants discussed communication. How can departments share more advanced data collection and data analysis? Some options
included using BlackBerries and MP3s.
The group also addressed the need for resources, especially with current budget cuts. However, the focus should be on outcomes.
Is the data being measured the most effective in terms of performance? The real value is to do more with existing or diminishing
In Steve Mastrofski and Shellie Solomon’s group, participants noted that good crime analysts are already practicing predictive
policing. Marginal improvements can be made and are areas of opportunity. The group highlighted the gap within agencies between
crime analysts and management. Analysts gather information and make recommendations to department administrators, but those
recommendations do not make it to the street-level cop.
A key point raised was that predictive policing should be community-based. Community leaders and the public must accept the
concept for it to be successful.
Additionally, departments must hire individuals trained to care about data. To succeed, predictive policing involves changing
the culture of policing by placing a higher value on information and its use.
Commander Jonathan Lewin provided the summary for the group that was also facilitated by Ed McGarrell and Captain Gary Lee.
Members asked whether they could ethically pilot predictive policing in just one area. If so, a cultural shift is necessary
in some organizations. He said that there was a need for accountability within departments to ensure success.
The group also discussed the need for data sharing and interoperable systems since crime does not respect agency boundaries.
To this end, there is a need to incorporate nontraditional data, including demographics and building foreclosures for more
Lewin explained how analysis is done in Chicago and how a cultural shift has occurred there. Every officer is considered a
crime analyst for his or her beat. Predictive policing will not replace the human component, but it would enhance that effort.
Further, the new data sets and tools of predictive policing won’t replace crime analysts but will assist them. There will
always need to be a human component. But the group acknowledged that there is too much data for one person to manage, so these
tools are necessary.
Second Predictive Policing Symposium
At the Second Predictive Policing Symposium, participants noted the importance of building a relationship with the community
and working with crime analysts. They recognized that mental health, probation, parole, corrections and other law enforcement
agencies should be involved in the predictive policing discussion. Other potential collaborators include police academies,
universities, fusion centers, community agencies, faith-based initiatives, public service agencies, policymakers, nonprofit
organizations, crime labs, medical examiners' offices and the media.
Date Modified: January 6, 2012